Subject: Re: [boost] Review Request : Boost.Range Extension
From: Michel Morin (mimomorin_at_[hidden])
Date: 2012-06-06 10:41:28
Akira Takahashi wrote:
>> IIUC, the output of `regular` does not satisfy the Regular concept.
>> Is this just an implementation deficiency or intentional behavior?
>> If it is intentional, then the name `regular` might not be a good name...
> Now fixing missing feature of Regular concept.
A few comments:
> template <class F>
> RegularFunctorType regular(F f)
* What is the requirement of `F`? CopyConstructible?
This should be documented, too.
* If `F` has relational operators, do `RegularFunctorType`'s relational
operators use them? If not, I think it would be better to document it.
* `indirect_functor` does not support `boost::result_of` protocol
for nullary function calls. See ticket #6914 for details
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk